Career track

The Data Science and ML track in India, and how it typically plays out.

DS, MLE, and Research Scientist are three related but distinct careers. This is how each one usually progresses at Indian and global product companies, and where the split happens.

Two caveats before you read further. First, "Data Scientist" is one of the most abused titles in Indian job listings. Read the JD carefully because the same title can mean anything from a SQL-only analyst to a research-heavy modeler. Second, the Data Scientist and ML Engineer ladders diverge early. Deciding which one you are on is the single most important career call in this space, and doing both well is much rarer than people claim.

The ladder, level by level

1. Data Analyst

Entry level. Business questions in, SQL and dashboards out.

Who ends up here: Business, statistics, math, and economics grads. Also engineering grads who lean toward business context over systems.
Typical day: Fields ad-hoc requests from PMs, builds recurring dashboards, cleans and joins data from multiple sources.
Signal for the next level: Owning a metric end to end, running your first A/B test, being trusted to answer strategy questions without a senior double-checking.

2. Data Scientist I / Junior DS

First real DS role. Owns a small experimentation surface or a set of models under supervision.

Who ends up here: Analysts with two to three years who moved up, or Masters grads with a portfolio of applied projects.
Typical day: Builds and evaluates models, designs and analyzes experiments, presents findings to the PM.
Signal for the next level: Framing your own problem statement instead of executing someone else's, and shipping a model that a product surface actually depends on.

3. Senior Data Scientist

Owns a full problem area end to end, from framing to deployed model. Mentors juniors.

Who ends up here: DS with four to seven years, ideally with prior deployment and monitoring experience, not just Jupyter notebooks.
Typical day: Splits time between modeling, cross-functional stakeholder conversations, and code review for junior DS.
Signal for the next level: Cross-team influence. Other teams start using your models or asking for your input on their modeling choices.

4. Staff / Lead DS

Sets modeling direction for a business unit or set of related products.

Who ends up here: Seven to twelve years of consistent DS delivery, with at least one deep specialization (recommendations, forecasting, causal inference, LLMs).
Typical day: Design docs, cross-team roadmaps, unblocking Senior DS, occasionally hands-on modeling on the hardest problem.
Signal for the next level: Company-wide impact. Systems you designed span multiple business units.

5. Principal DS

Company-wide DS direction. Very rare below Series C or public-company scale.

Who ends up here: Twelve plus years. Deep expertise plus a track record of framing problems that others would not have seen.
Typical day: Mostly strategy, mostly writing, occasional deep-dive modeling on the most consequential decisions.
Reality check: Very few Principal DS roles exist in India today. If this is your goal, be strategic about the size and stage of the companies you join.

6. ML Engineer track (parallel)

MLE-1 to Senior MLE to Staff. Focus on serving, features, monitoring, evaluation.

When to consider it: If you find yourself enjoying the systems side more than the modeling side. Ownership of production ML infra is currently one of the most compensated specializations in India.
Compensation: Broadly parallel to the DS ladder, often 10 to 20 percent higher at the same level because supply is tighter.
Overlap with SDE: Growing. Many MLE roles today are effectively backend engineer roles with model-serving specialization.

7. Research Scientist (separate track)

PhD-heavy roles at labs. Different hiring bar, different day-to-day.

Where it exists: Google Research India, Microsoft Research, Adobe Research, Amazon Science, IBM Research India, IISc-affiliated labs, a handful of GenAI-native startups.
What it involves: Publishing, applied research, occasional productionization. Slower pace, longer horizons, higher intellectual freedom.
Realistic entry: Almost always requires a PhD in ML or a closely related field, or an unusually strong Masters with published work.

Approximate salary ranges

Cash compensation at tier-1 product companies in India. Stock and bonus on top and vary widely by company stage. Research-lab and specialist LLM roles can be significantly higher.

Level	Range (LPA)
Data Analyst	8 to 15 LPA
Data Scientist I	18 to 32 LPA
Senior Data Scientist	40 to 70 LPA
Staff / Lead DS	65 to 110 LPA
Principal DS	100 to 170 LPA
ML Engineer II	35 to 60 LPA
Senior ML Engineer	55 to 90 LPA
Research Scientist (labs)	60 to 180 LPA, highly variable

Common pivots

DS to PM

Extremely common at product companies. Senior DS with strong business framing often move into PM roles owning data or ML-heavy products.

DS to MLE

Requires a real investment in backend engineering, systems, and cloud infra. Doable within a year of focused study if you were already technical.

DS to Founder

GenAI-native startups have made this pivot popular. Works best if you have a domain-specific insight that a wrapper company cannot easily replicate.

Analyst to Growth or Strategy PM

For analysts who love business context more than modeling. Growth or Strategy PM roles at consumer companies are a natural landing.

Let Hiro pick the DS and MLE roles worth your Saturday.

Tell Hiro where you are on the DS or MLE ladder, the domain you want to specialize in, and the stage of company that fits. Every Saturday morning you get a short list of 5 roles with an honest note on each.

Meet Hiro

Common questions

Is Data Scientist the same as ML Engineer?

Not really, though a lot of Indian job descriptions use them interchangeably. A Data Scientist usually spends most of their time on problem framing, analysis, experimentation, and modeling for insight. An ML Engineer usually spends most of their time productionizing models, building serving infrastructure, and owning the pipeline that keeps models fresh. If you love statistics and business context, the DS side fits. If you love systems and infrastructure, the MLE side fits. Some rare people do both well but that combination is unusual.

Is a PhD required for a Data Science role in India?

For applied DS at product companies, no. A strong Masters or a very strong Bachelors with real projects is enough for most Senior DS roles. For research-heavy roles at labs (Google Research India, Microsoft Research, Adobe Research, Amazon Science, some IISc-affiliated labs) a PhD helps significantly. For everything in between, portfolio and reasoning ability matter more than the degree.

How do I break into DS from a non-CS background?

Business, math, statistics, economics, and physics grads move into DS regularly. The most reliable path is Analyst to Senior Analyst inside a product or fintech company, then formal migration to a DS role once you have owned an experimentation or business insight track for at least a year. Bootcamps are useful for foundation but rarely land you a Senior DS role by themselves.

What is a realistic Senior DS salary in India?

For a Senior Data Scientist with 4 to 7 years at a tier-1 product company, cash compensation typically lands between 40 and 70 LPA. Applied ML Engineers at the same level often see 5 to 15 LPA more because supply is tighter. Research roles at labs like Google or Microsoft can be substantially higher, but count as a separate market.

How is generative AI changing the DS career path?

The demand for people who can fine-tune, evaluate, and productionize LLMs is significantly outpacing supply. Engineers who invested a year in serious LLM work over the last 24 months have skipped one or two levels of the ladder in real terms. That window is narrowing. By 2027 the pattern will normalize and LLM literacy will be table stakes for most DS and MLE roles, not a differentiator.